Yeah, I totally see where you're coming from, though consider that finding the riggedness is interesting enough:
How likely is that they would have the time / patience / skill to manipulate the data to actually be statistically legitimate? Depending on the dataset, a quick check for Benford's law would find said riggedness.
It's not that the data is illegitimate per se, but it makes any conclusions from the data unreliable. (e.g. "When is the best time to post on Product Hunt?" is difficult since the raw data alone does not account for the impact of having an influencer post the submission, and asking for upvotes through networks)
